reported similar findings in boys, based on a standardized assessment of maturation by physicians who examined the boys at age 15 (35). Family Size and Intelligence. We have seen that several health indicators reflect the effects of family size. Another important indicator, although perhaps of well-being rather than health per se, is intelligence. Intelligence has been extensively studied in relation to family size, and the existence of a striking negative correlation between the two has been known for many years. Anastasi reviewed the literature carefully a decade ago (41); Hunt has discussed it thoughtfully more recently (42). Studies of large populations have shown repeatedly that children from large families score significantly lower in intelligence tests. For example, the results of a group intelligence test administered to most of the 11-year-olds in Scotland in 1932 and in 1947 were analyzed by the Scottish Council for Research in Education. In 1932 the sample numbered 87,498, and in 1947 it was 70,805. This represented 87 and 88 percent of the total populations, respectively. In both studies, a negative correlation was found between the test scores and size of sibship (number of children in a family). This negative correlation held true in all social classes, even though the children's scores reflected social class differences. A random sample of 1,215 of the children tested in 1947 were given the Stanford Binet intelligence test; again the negative correlation with sib-ship was found.* The average I.Q. of only children was 113; that of children with five siblings or more was 91. A second large-scale study reviewed by Anastasi was carried out in France during 1943-44, when 2 percent of the total elementary school population of France, age 6 to 12, was tested (41). The findings were almost identical: mean test scores decreased with increase in family size. Only children were found to have an average mental age 1 to 2 years higher than children with eight siblings or more, and the differential was apparent in each age group tested. Finally, this study showed the expected social class differences in scores, but the negative correlation between intelligence and family size varied among the classes. It was "clearly apparent" in children of fanners, manual laborers, and clerical workers, "negligible" in children from the managerial class, and "barely discernible" among those from the professional class. Anastasi also cites numerous other studies of normal children carried out in England, the United States, Greece, and Germany in which it was found that mean intelligence test scores declined with increase in family size. In addition, she notes that Terman, in his study of gifted children, found a negative correlation' between I.Q. and number of children in a family (41). *In this case the correlation coefficient was -.32. 1> = -.27.